To improve forecasting accuracy of the generated energy of regional small hydropower stations, it is proposed a forecasting method of regional small hydropower generation considering the cumulative effect of rainfall.In this paper, the meteorological data was classified by using decision tree firstly. Then through the in-depth analysis of the output law of small hydropower stations, the generated energy forecasting model was set up using the method of multiple regression fitting that considering the rainfall effect. Finally, the corrected model is established to correct the predicted results that considering accumulative effect of rainfall.
INTRODUCTIONAs an important clean renewable energy, small hydroelectric power plants have following features. The installed capacity is small, the construction period is short, the output is closely related to weather, and they are widespread throughout the area with abundant water resources.With the installment capacity of small hydroelectric power plants growing, its proportion in total electrical generating capacity is rising. Due to the small reservoir capacity and no energy storage, rainfall greatly influences the generating capacity of small hydroelectric power plants, that is, the generating capacity of small hydroelectric power plants depends on the rainfall and then the wholesale electricity is affected indirectly.A great deal of research work on the influence of meteorological factors on the load forecasting curve has been conducted by scholars around the world[1-6], but there are still several aspects that should be considered.(1) The complexity and variability of weather ,including the weather changes continuously in time,and the weather might abruptly change at the same time.(2) Dual nature of the influence of weather on load. One hand, the effect of real-time weather changes on the load. On the other hand, the effect of meteorological cumulative effect on the load.For the sake of those problems mentioned above, a forecasting method based on meteorological decision tree was proposed ,and the monthly generated energy of small hydropower stations was predicted by using the method in this paper.